The standard NFT explainer treats Bitcoin NFTs and Ethereum NFTs as the same thing on different chains. They aren't. The architectural differences run deeper than which validators secure the chain. They go to the underlying accounting model, the storage approach, the security guarantees, and the marketplace economics. A serious collector or developer needs to understand these differences because they change what's possible and what's safe.

This piece walks through the structural differences honestly. Some of them favor Bitcoin. Some favor Ethereum. The point is to understand the differences, not to crown a winner.

The accounting model: UTXO vs account

Bitcoin uses the Unspent Transaction Output (UTXO) model. Every Bitcoin transaction consumes one or more existing UTXOs as inputs and produces one or more new UTXOs as outputs. Your "balance" isn't a number stored anywhere; it's the sum of all UTXOs that you control private keys for. There's no global Bitcoin state showing each address's balance; there's only the chain of transactions and the UTXO set derived from them.

Ethereum uses the account model. Every address has a balance, stored explicitly in the chain's state. A transaction debits one account and credits another. Smart contracts are accounts with code that runs when invoked.

This difference shapes everything downstream.

On Bitcoin, an NFT is a UTXO (Ordinals: a specific satoshi within a UTXO; Counterparty: a token balance associated with an address that's encoded in transaction history; Stamps: data embedded in UTXO outputs). The non-fungible token is an on-chain object that exists in the chain's state because the UTXO set requires it. Holding the NFT means controlling the keys to the UTXO.

On Ethereum, an NFT is an entry in a smart contract's state, typically following the ERC-721 standard. The contract maintains a mapping from token IDs to owner addresses. The "token" doesn't have an independent existence; it's a database row inside a smart contract.

Storage: on-chain content vs off-chain references

Bitcoin protocols generally put more content directly on-chain than Ethereum protocols do.

An Ordinals inscription stores the image, video, or text bytes themselves in Bitcoin's witness data. A Bitcoin Stamp stores image bytes in UTXOs. A Counterparty asset stores metadata on-chain with the image off-chain (the Rare Pepe Directory model). But even the off-chain Counterparty model is fundamentally different from typical Ethereum NFT practice.

Most Ethereum ERC-721 NFTs are pointers. The smart contract holds a token URI for each token, and that URI points to a JSON file (usually on IPFS or HTTPS) which itself points to the actual image asset (usually on IPFS). The chain holds a pointer to a pointer to the data. If the IPFS pin lapses or the centralized server goes down, the visual representation of the NFT can become unavailable even though the on-chain token entry persists.

The trade-offs are real. Ethereum's approach is cheaper per token but introduces external dependencies. Bitcoin's approaches (Ordinals especially) are more expensive but reduce the dependency chain. Counterparty sits in between, with on-chain metadata and off-chain content that has nonetheless survived continuously since 2014.

For collectors thinking about long-term durability, the Bitcoin-side approaches are architecturally better positioned for multi-decade survival of the visual asset itself. For high-volume, low-cost issuance, the Ethereum approach is more efficient. This isn't a Bitcoin-is-better claim; it's a design-choice difference.

Smart contracts: there aren't any on Bitcoin

Ethereum's NFT marketplaces, royalty enforcement, batch transfers, and complex auction mechanisms all run as smart contracts. The contract holds state, executes code in response to transactions, and enables features that simply don't exist on Bitcoin in the same form.

Bitcoin doesn't have smart contracts in the Ethereum sense. Bitcoin Script is a stack-based language with intentionally limited expressiveness; you can't write arbitrary state machines in it. What Bitcoin has instead is a set of conventions on top of Bitcoin Script (multisig, time-locks, hash-locks, P2SH/P2WSH wrapping) that enable some contract-like behavior.

The implications for NFTs:

  • No on-chain royalty enforcement. When a Bitcoin NFT changes hands, the protocol has no mechanism to redirect a portion of the sale price to the original creator. Royalties are entirely a marketplace-side convention; if the marketplace doesn't enforce them, royalties don't get paid. This is true of Ordinals, Counterparty, and Stamps alike.
  • Trading happens via atomic swaps or marketplace contracts. The Counterparty DEX has been doing atomic swaps since 2014. Modern Ordinals marketplaces use partially-signed Bitcoin transactions (PSBTs) to enable trustless trade settlement.
  • No "approve and forget" pattern. Ethereum users routinely grant unlimited approvals to marketplace contracts (which has led to many exploits). Bitcoin NFT trading patterns rely more on per-transaction signing or atomic swaps that don't require persistent approvals.

The royalty question

Ethereum NFT royalties were never enforced at the protocol level either; ERC-721 doesn't specify royalties. The expectation that 5-10% of secondary sales would flow to creators was a marketplace convention enforced by OpenSea and other major venues. When Blur and other newer marketplaces decided not to enforce royalties in 2022-2023, the convention broke down.

Bitcoin NFTs inherited this state. Most Bitcoin marketplaces don't enforce royalties. Some Ordinals projects have experimented with workarounds (artificial scarcity created through inscription-burning, off-chain agreements with marketplaces), but the absence of contract-level royalty enforcement means the economic model for creators is different.

For artists, this is a real consideration. The model that works is the Counterparty-era one: low-or-no commission on primary sales, scarcity managed at issuance, value accruing to the creator on the primary mint rather than via secondary royalties. The 2021-era ERC-721 royalty assumption never reliably worked even on Ethereum, and it doesn't work on Bitcoin either.

Security model differences

An Ethereum NFT smart contract can have bugs. If the contract is upgradeable, the upgrade authority can change behavior. If the contract has admin functions, those functions can be exploited. The history of Ethereum NFTs includes contract-level exploits that drained collections and admin-key compromises that disabled trading.

A Bitcoin NFT has different vulnerabilities. Counterparty assets depend on the Counterparty protocol layer; a bug in Counterparty itself could theoretically affect asset behavior, though the protocol has been remarkably stable over 12 years. Ordinals inscriptions depend on the ord indexer software interpreting witness data correctly; different indexers could in principle disagree, and a contested interpretation has historically been resolved in favor of the dominant indexer's reading. Stamps depend on the consensus around how UTXO-embedded data should be parsed.

The architectural difference: Bitcoin NFTs inherit Bitcoin's security model directly (the asset is a UTXO, Bitcoin secures the UTXO set), but they're vulnerable to bugs in the metaprotocol or indexer layer. Ethereum NFTs inherit Ethereum's security model plus the security of the specific contract, which is variable.

Marketplace economics

The biggest Ethereum NFT marketplaces (OpenSea, Blur, the late LooksRare) have all charged trading fees and competed for volume by adjusting royalty enforcement. The economics are fundamentally similar to centralized exchange economics: the marketplace is a venue with margin.

Bitcoin NFT marketplaces operate in a more fragmented market with weaker network effects. The recent Magic Eden exit redistributed approximately 80% of Bitcoin Ordinals trading volume across multiple smaller platforms. Counterparty trading has always been distributed across the original Rare Pepe Wallet, the Counterparty DEX, and modern multi-protocol marketplaces. The economics are closer to a federation of marketplaces than a winner-take-all platform.

For collectors, this means more comparison shopping and less platform lock-in. For marketplaces, it means harder unit economics. For the broader category, it means innovation in trade settlement (trustless atomic swaps, on-chain DEX) tends to come from Bitcoin-side platforms with no easy way to extract custodial rents.

What this all means

A Bitcoin nft behaves differently from its Ethereum equivalent in ways that aren't visible on the surface. Different storage. Different security model. Different royalty economics. Different marketplace structure. The visual asset might look similar; the underlying object is fundamentally distinct.

Neither approach is universally better. Ethereum NFTs offer more contract-level flexibility, richer royalty mechanisms (when enforced), and lower per-token costs. Bitcoin NFTs offer stronger architectural guarantees about asset persistence, no smart-contract bug surface, and longer track records (Counterparty has been operating for over a decade longer than any Ethereum NFT contract).

For collectors making genuine architectural choices: hold both. The protocols solve different problems. The reasonable position is that Bitcoin NFTs and Ethereum NFTs are complementary technologies, each better-suited to certain use cases, and the binary "which is better" framing obscures more than it reveals.